Sixense VR SDK

In the end something other than standard Vibe Motors would be nice(Waiting for a Haptic Feedback company to write back to me atm). So Ill let a bit of the cat out of the bag. What I am making is trying to be an "affordable" VR Interaction Gun( I won't draw a comparison but we know who that is), and I will be creating some sort of site so that you might follow the progress. just remember its really just me, so progress is slow, I do have 'some' help but not enough, we are looking for a programmer if anyone is interested. Also I would LOVE to hear peoples ideas, so if you have an interesting idea for rumble controllers, or an interesting tracking scheme? Right now the gun house a 9DOF IMU, and only relative mouse movements atm, but I think I will get a video together....good idea;O)

"lmaceleighton" wrote:
In the end something other than standard Vibe Motors would be nice(Waiting for a Haptic Feedback company to write back to me atm).

A solenoid actuator (electromagnetic coil, metal rod and a spring) aligned with the barrel of the gun would give a better approximation of a shot than a vibration motor. The old arcade game Gunblade worked like that (I believe). A sliding bar inside it's machine guns gave a 1 dimensional feedback and made a lot of noise. My arms felt sore if I played for too long. :)
Although I'm not sure what the power requirements of one strong enough for gun feedback would be.

I have been looking into that, and I think that is similar to what the guys at Delta Six are doing, and that is a great solution, so if I have the room then I will go with something like this for the final version. Right now I have two standard rumble motor taken right out of a standard Playstation 2 controller. I was thinking of going to kickstarter once the initial prototype is finished. I was trying not be competition for Delta Six,as I feel that the product I am trying to make differs in a few fundamental ways. One thing that I am trying very hard to do is to keep it AFFORDABLE. I don't want this thing to coast over $150, and right now that is looking VERY possible. Another difference is that I am tailoring this VR Gun to work with the Oculus Rift, it is the #1 goal, as our final product is supposed to be the gun AND a game to go with it designed specifically for the rift. Also I feel like I have hijacked this thread now, so I am going to start my own, just to be polite;O)

So let do this all here:

viewtopic.php?f=25&t=8204

Hi Owen,

Two-way physics simulation is indeed tricky. Especially because your physical body/hands may get out of sync with the virtual body (try slamming your hand against a virtual wall).
I see from your 'Armored Ops' video you have been dealing with this. What is your approach to this?

In my case I will let the virtual and physical hands disconnect when virtual physics demand it. In the case of the wall, the virtual hand stops while the physical hand goes on. It is my experience that this 'feels' much better than letting the virtual hand penetrate the wall. Besides that, you still have the natural reflex to stop your physical hand when it looks like it will be hitting a wall.

Are you also considering applying forces to the environment? Moving heavy objects should have a different effect than moving light objects.

I saw you can walk around. Is this by animation or using procedural walking? I am making my first steps with the latter now ;-)

PS: somehow the email notifications on threads did not work for me. I thought the thread was dead...

My observations were similar, but I have not settled on a particular way of implementing this for SixenseVR yet, so nothing concrete to announce. Walking is animation driven, but I think there is definitely a future for procedural approaches.

Armored Ops used simulated motors with torque limits. This provided all the behaviors you describe but was fairly complicated to set up and had some inherent acceleration latency. I am experimenting around these problems and hope to have something fairly robust, but precision and predictability are top priority for launch.

Of course quality above quantity (of features)! But exploring the new domain and trying to solve these new problems is very tempting...
I will be preparing a new video on my progress next week and possibly also a free preview download of my work. I am already looking forward to the comments.

I will be following you progress with great interest, Owen!

I'm sad that I'm a month late to the convo, but I just wanted to mention that I've experimented with motorized joints and IK (in 2D).

http://www.openprocessing.org/sketch/84256

Owen's right about the acceleration latency... but perhaps because of the IMU that's been added to the system, you can apply direct accelerations (from the accelerometer) to the arm rig in addition to the motorized joints to mitigate the latency without giving the arms infinite strength. Scale the acceleration so it's just enough to offset the motorized joint latency...

Here is the video I mentioned earlier:



You can download the package at here: https://serrarens.nl/passervr/downloads/vr-hands/